An algorithm with reduced computational complexity for estimation of parameters of linear stationary discrete systems

The estimation of linear stationary discrete systems parameters (LSDS) by the measured input and output values is of major importance in control systems synthesis. Most of the methods for parameters estimation are based on the least squares method which implementation is related to matrices inversion. The present paper aims to propose an algorithm for estimation of parameters of LSDS, described in the state-space, in which, the size of the inverted matrices is always n×n. Proposed is a procedure for matrices inversion with reduced computational complexity, by using the least squares method (LSM).